- Pacman-dev - lists.archlinux.org

[pacman-dev] [PATCH 4/5] check for overflow when setting HTTP_USER_AGENT
by Andrew Gregory 10 May '17

10 May '17

gcc7 issues a warning about a potential overflow if left unchecked. Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- src/pacman/pacman.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/src/pacman/pacman.c b/src/pacman/pacman.c index 605aec3e..11b7e6f0 100644 --- a/src/pacman/pacman.c +++ b/src/pacman/pacman.c @@ -271,10 +271,15 @@ static void setuseragent(void) { char agent[101]; struct utsname un; + int len; uname(&un); - snprintf(agent, 100, "pacman/%s (%s %s) libalpm/%s", + len = snprintf(agent, 100, "pacman/%s (%s %s) libalpm/%s", PACKAGE_VERSION, un.sysname, un.machine, alpm_version()); + if(len >= 100) { + pm_printf(ALPM_LOG_WARNING, _("HTTP_USER_AGENT truncated\n")); + } + setenv("HTTP_USER_AGENT", agent, 0); } -- 2.12.2

1 0

[pacman-dev] [PATCH 3/5] alpm_list: abort on memory allocation failure
by Andrew Gregory 10 May '17

10 May '17

This makes it possible to detect a failure in several alpm_list functions. Previously these functions would continue after a failure, returning partial results and potentially leaking memory. Unfortunately, NULL is a valid return value for the affected functions if the input list is empty, so they still do not have a dedicated error value. Callers can at least detect an error by checking if the input list was empty. Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- There are more functions that fail silently, but they can't be fixed without breaking API. lib/libalpm/alpm_list.c | 29 ++++++++++++++++++++++++----- 1 file changed, 24 insertions(+), 5 deletions(-) diff --git a/lib/libalpm/alpm_list.c b/lib/libalpm/alpm_list.c index 0f1b819c..334f67ba 100644 --- a/lib/libalpm/alpm_list.c +++ b/lib/libalpm/alpm_list.c @@ -461,7 +461,10 @@ alpm_list_t SYMEXPORT *alpm_list_remove_dupes(const alpm_list_t *list) alpm_list_t *newlist = NULL; while(lp) { if(!alpm_list_find_ptr(newlist, lp->data)) { - newlist = alpm_list_add(newlist, lp->data); + if(alpm_list_append(&newlist, lp->data) == NULL) { + alpm_list_free(newlist); + return NULL; + } } lp = lp->next; } @@ -480,7 +483,10 @@ alpm_list_t SYMEXPORT *alpm_list_strdup(const alpm_list_t *list) const alpm_list_t *lp = list; alpm_list_t *newlist = NULL; while(lp) { - newlist = alpm_list_add(newlist, strdup(lp->data)); + if(alpm_list_append_strdup(&newlist, lp->data) == NULL) { + FREELIST(newlist); + return NULL; + } lp = lp->next; } return newlist; @@ -498,7 +504,10 @@ alpm_list_t SYMEXPORT *alpm_list_copy(const alpm_list_t *list) const alpm_list_t *lp = list; alpm_list_t *newlist = NULL; while(lp) { - newlist = alpm_list_add(newlist, lp->data); + if(alpm_list_append(&newlist, lp->data) == NULL) { + alpm_list_free(newlist); + return NULL; + } lp = lp->next; } return newlist; @@ -523,8 +532,15 @@ alpm_list_t SYMEXPORT *alpm_list_copy_data(const alpm_list_t *list, void *newdata = malloc(size); if(newdata) { memcpy(newdata, lp->data, size); - newlist = alpm_list_add(newlist, newdata); + if(alpm_list_append(&newlist, newdata) == NULL) { + free(newdata); + FREELIST(newlist); + return NULL; + } lp = lp->next; + } else { + FREELIST(newlist); + return NULL; } } return newlist; @@ -552,7 +568,10 @@ alpm_list_t SYMEXPORT *alpm_list_reverse(alpm_list_t *list) list->prev = NULL; while(lp) { - newlist = alpm_list_add(newlist, lp->data); + if(alpm_list_append(&newlist, lp->data) == NULL) { + alpm_list_free(newlist); + return NULL; + } lp = lp->prev; } list->prev = backup; /* restore tail pointer */ -- 2.12.2

1 0

[pacman-dev] [PATCH 2/5] query_fileowner: avoid buffer overflow
by Andrew Gregory 10 May '17

10 May '17

Copying a string into a buffer that has just been determined to not be able to hold it is obviously incorrect. The actual error handling appears to have been unintentionally removed in 47762ab687959e48acc2de8592fcf3ba3cfa502b. Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- src/pacman/query.c | 1 + 1 file changed, 1 insertion(+) diff --git a/src/pacman/query.c b/src/pacman/query.c index 119764bc..247423fa 100644 --- a/src/pacman/query.c +++ b/src/pacman/query.c @@ -205,6 +205,7 @@ static int query_fileowner(alpm_list_t *targets) size_t rlen = strlen(rpath); if(rlen + 2 >= PATH_MAX) { pm_printf(ALPM_LOG_ERROR, _("path too long: %s/\n"), rpath); + goto targcleanup; } strcat(rpath + rlen, "/"); } -- 2.12.2

1 0

[pacman-dev] [PATCH 1/5] alpm_unlock: check handle before dereferencing
by Andrew Gregory 10 May '17

10 May '17

Prevents SIGSEGV if we catch SIGINT or SIGHUP before initializing alpm. Signed-off-by: Andrew Gregory <andrew.gregory.8(a)gmail.com> --- lib/libalpm/handle.c | 1 + 1 file changed, 1 insertion(+) diff --git a/lib/libalpm/handle.c b/lib/libalpm/handle.c index 502a5d9e..08625e3a 100644 --- a/lib/libalpm/handle.c +++ b/lib/libalpm/handle.c @@ -135,6 +135,7 @@ int _alpm_handle_lock(alpm_handle_t *handle) */ int SYMEXPORT alpm_unlock(alpm_handle_t *handle) { + ASSERT(handle != NULL, return -1); ASSERT(handle->lockfile != NULL, return 0); ASSERT(handle->lockfd >= 0, return 0); -- 2.12.2

1 0

[pacman-dev] [PATCH] Handle empty string passed to query_owner
by Allan McRae 10 May '17

10 May '17

Passing an empty string to pacman -Qo results in: error: No package owns <first directory in $PATH> Treat an empty string being passed the same as recieving a NULL value and exit searching for an owner. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- Running 'pacman -Qo ""' will now just exit pacman. I'm not sure if we need an error/warning message for this case. src/pacman/query.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/src/pacman/query.c b/src/pacman/query.c index a8417570..119764bc 100644 --- a/src/pacman/query.c +++ b/src/pacman/query.c @@ -161,7 +161,8 @@ static int query_fileowner(alpm_list_t *targets) size_t len, is_dir; unsigned int found = 0; - if((filename = strdup(t->data)) == NULL) { + filename = strdup(t->data); + if(filename == NULL || strcmp(filename, "") == 0) { goto targcleanup; } -- 2.12.2

2 1

[pacman-dev] [PATCH] Do not continuously try to open an invalid database
by Allan McRae 10 May '17

10 May '17

If you manage to download a bad database (e.g. an html file when behind a proxy or with a badly configured webserver), pacman makes sure you know about it. Here is some example output: error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format error: could not open file /var/lib/pacman/sync/extra.db: Unrecognized archive format I don't know how many times that gets printed because it goes beyond my scrollback buffer. Flag a database that we can "open" and "fstat" but not read from as invalid to avoid this. Signed-off-by: Allan McRae <allan(a)archlinux.org> --- lib/libalpm/be_sync.c | 3 +++ lib/libalpm/util.c | 1 + 2 files changed, 4 insertions(+) diff --git a/lib/libalpm/be_sync.c b/lib/libalpm/be_sync.c index 06f509a6..46959298 100644 --- a/lib/libalpm/be_sync.c +++ b/lib/libalpm/be_sync.c @@ -486,6 +486,9 @@ static int sync_db_populate(alpm_db_t *db) fd = _alpm_open_archive(db->handle, dbpath, &buf, &archive, ALPM_ERR_DB_OPEN); if(fd < 0) { + if(db->handle->pm_errno == ALPM_ERR_DB_INVALID) { + db->status &= DB_STATUS_INVALID; + } return -1; } est_count = estimate_package_count(&buf, archive); diff --git a/lib/libalpm/util.c b/lib/libalpm/util.c index 35fc7f41..0fb9bb6d 100644 --- a/lib/libalpm/util.c +++ b/lib/libalpm/util.c @@ -268,6 +268,7 @@ int _alpm_open_archive(alpm_handle_t *handle, const char *path, if(archive_read_open_fd(*archive, fd, bufsize) != ARCHIVE_OK) { _alpm_log(handle, ALPM_LOG_ERROR, _("could not open file %s: %s\n"), path, archive_error_string(*archive)); + error = ALPM_ERR_DB_INVALID; goto error; } -- 2.12.2

2 1

[pacman-dev] bug in 'who owns' feature?
by Adesh Kumar 07 May '17

07 May '17

Tried using Pacman v5.0.1 - libalpm v10.0.1 --($:~)-- pacman -Qo "" error: No package owns /usr/local/sbin/ --($:~)-- pacman -Qo /usr/local/sbin/ /usr/local/sbin/ is owned by filesystem 2016.12-2

1 0

[pacman-dev] pacman sorting incompatible with 'sort' command
by Adesh Kumar 05 May '17

05 May '17

pacman internal sorting on package names differ from the one implemented in linux 'sort' command. pacman treats '-' (hyphen) differently. Testcase: 1. Install i3 group. It will install i3-wm, i3lock and i3status 2. Check the following commands: --($:~)-- pacman -Qeq | grep i3 i3-wm i3lock i3status --($:~)-- pacman -Qeq | grep i3 | sort i3lock i3status i3-wm --($:~)-- pacman -Qeq | grep i3 | sort -c sort: -:2: disorder: i3lock

5 4

Re: [pacman-dev] [PATCH 3/4] makepkg: unify source file times for improved build reproducibility
by Allan McRae 02 May '17

02 May '17

On 21/04/17 13:36, Eli Schwartz wrote: > On 04/20/2017 11:01 PM, Allan McRae wrote: >> I am probably moving this to after source extraction/prepare() running, >> so it can be skipped with --noextract. > > But --noextract depends on your having at some point previously run > --nobuild, in order to pull updated sources, re-patch any patches, etc. > which I still don't want to do manually outside of makepkg... I don't > see why makepkg should start breaking things for me. > > How about instead we guard it with > BUILDENV+=(fix_everyone_elses_SOURCE_DATE_EPOCH_stuff) > > Or better yet, just file bugs against whatever upstream build > system/programming language/source code is determined to sneakily embed > source code modification times into generated files, and call it a day? > Adding list back - any further off-list replies will be completely ignored. The reproducible builds people will provide details, but it appears pyo/pyc do this. A

3 3

Re: [pacman-dev] [RFC] Make PKGBUILD attributes configurable
by Allan McRae 24 Apr '17

24 Apr '17

On 23/04/17 09:36, Dustin Falgout wrote: > I would like a way to include custom attributes from the PKGBUILD in the output of the --printsrcinfo option. So basically, this... > > pkgbase = pacman > pkgdesc = A library-based package manager with dependency support > pkgver = 5.0.1 > pkgrel = 4 > url = http://www.archlinux.org/pacman/ > arch = i686 > arch = x86_64 > ... > _custom_attribute1 = some value > _custom_attribute2 = some value > _custom_attribute3 = some value > ... > > pkgname = pacman > Bringing the mailing list back into this (seems our reply field is broken...) Can you give an example of what a custom attribute would be? Your example is still to vague to judge whether this would be something we wish to support. Thanks, Allan

2 3